Constrained color dithering

Abstract

In one embodiment, a computing system may receive a target image to be displayed on a display. The target image may have a first number of bits per color. The system may access a mask for each color channel of RGB color channels. The accessed masks may be generated based on (1) one or more characteristics of a human visual system, and (2) a subpixel geometry layout of the display. The system may generate an output image based on the target image and the accessed masks. The output image may have a second number of bits per color smaller than the first number of bits per color. The system may display the output image on the display for representing the target image.

What is claimed is: 
     
         1 . A method comprising, by a computing system:
 receiving a target image to be displayed on a display, the target image having a first number of bits per color;   accessing a mask for each color channel of RGB color channels, wherein the accessed masks are generated based on (1) one or more characteristics of a human visual system, and (2) a subpixel geometry layout of the display;   generating an output image based on the target image and the accessed masks, wherein the output image has a second number of bits per color smaller than the first number of bits per color; and   displaying the output image on the display for representing the target image.   
     
     
         2 . The method of  claim 1 , wherein the display has the second number of bits per color that is smaller than the first number bits per color of the target image. 
     
     
         3 . The method of  claim 1 , wherein the mask of that color channel comprises a plurality of dot patterns each corresponding to a grayscale level for that color channel, and wherein the plurality of dot patterns satisfies a spatial stacking property. 
     
     
         4 . The method of  claim 3 , wherein each dot pattern of the plurality of dot patterns comprises all dot patterns of lower grayscale levels of that color channel. 
     
     
         5 . The method of  claim 1 , wherein the masks for the RGB color channels satisfy a color stacking property, and wherein a dot that is assigned to a color remains that color when a target color becomes brighter along a color space path starting from black and ending at white. 
     
     
         6 . The method of  claim 1 , wherein the mask for that color channel is generated using a simulated annealing process comprising:
 generating a first dot pattern for a target grayscale level of that color channel;   calculating a first perceptual error for the first dot pattern;   generating a second dot pattern by swapping at least one dot in the first dot pattern to a different dot position;   calculating a perceptual error for the second dot pattern; and   determining whether to keep or discard the second dot pattern based on a comparison of the first perceptual error and the second perceptual error.   
     
     
         7 . The method of  claim 6 , wherein the first and second perceptual errors are calculated using a cost function, and wherein the cost function has a first constraint based on the one or more characteristics of the human visual system and a second constraint based the subpixel geometry layout of the display. 
     
     
         8 . The method of  claim 7 , wherein the one or more characteristics of the human visual system are represented by a matrix representing different sensitives of the human visual system at different chromatic channels or within different spatial frequency ranges 
     
     
         9 . The method of  claim 7 , wherein the subpixel geometric layout of the display is represented by a matrix representing sizes, shapes, and layouts of subpixels of the display. 
     
     
         10 . The method of  claim 7 , wherein the cost function iterates through a plurality of chromatic channels and all dots of a current dot pattern to determine a perceptual error value. 
     
     
         11 . The method of  claim 6 , wherein, when the display has a larger subpixel size in a current color channel, and wherein the at least one dot comprises a group of dots that are swapped together to represent to a subpixel with the larger subpixel size. 
     
     
         12 . The method of  claim 6 , wherein the at least one dot being swapped is randomly selected, and wherein the at least dot is swapped to the different dot position as constrained by the subpixel geometry layout of the display. 
     
     
         13 . The method of  claim 6 , wherein the mask generated using the simulated annealing process provides a minimized perceptual error value. 
     
     
         14 . One or more computer-readable non-transitory storage media embodying software that is operable when executed to:
 receive a target image to be displayed on a display, the target image having a first number of bits per color;   access a mask for each color channel of RGB color channels, wherein the accessed masks are generated based on (1) one or more characteristics of a human visual system, and (2) a subpixel geometry layout of the display;   generate an output image based on the target image and the accessed masks, wherein the output image has a second number of bits per color smaller than the first number of bits per color; and   display the output image on the display for representing the target image.   
     
     
         15 . The media of  claim 14 , wherein the display has the second number of bits per color that is smaller than the first number bits per color of the target image. 
     
     
         16 . The media of  claim 14 , wherein the mask of that color channel comprises a plurality of dot patterns each corresponding to a grayscale level for that color channel, and wherein the plurality of dot patterns satisfies a spatial stacking property. 
     
     
         17 . The media of  claim 16 , wherein each dot pattern of the plurality of dot patterns comprises all dot patterns of lower grayscale levels of that color channel. 
     
     
         18 . A system comprising:
 one or more non-transitory computer-readable storage media embodying instructions; and   one or more processors coupled to the storage media and operable to execute the instructions to:
 receive a target image to be displayed on a display, the target image having a first number of bits per color; 
 access a mask for each color channel of RGB color channels, wherein the accessed masks are generated based on (1) one or more characteristics of a human visual system, and (2) a subpixel geometry layout of the display; 
 generate an output image based on the target image and the accessed masks, wherein the output image has a second number of bits per color smaller than the first number of bits per color; and 
 display the output image on the display for representing the target image. 
   
     
     
         19 . The system of  claim 18 , wherein the display has the second number of bits per color that is smaller than the first number bits per color of the target image. 
     
     
         20 . The system of  claim 18 , wherein the mask of that color channel comprises a plurality of dot patterns each corresponding to a grayscale level for that color channel, and wherein the plurality of dot patterns satisfies a spatial stacking property.